Research On Fault Detection And Location Of Train Network

Train communication network is mainly composed of central control unit,train level network,vehicle level network,on-board equipment and transmission medium,which is the center of train operation control.With the intelligent development of high-speed trains in China,it is the general trend to increase network nodes such as passenger service and Internet of things.Therefore,the topological structure of train communication network will be more and more complex,and the research on fault detection and location of train communication network is more and more urgent.Train fault detection involves many components,which is a complex system engineering.At present,the research on fault detection of train communication network in China is still relatively backward,the research on fault location of train network transmission medium is rare,and the research on fault of on-board equipment only stays on hardware fault.Therefore,based on the national science and technology support project "key technology and system development of train control and information service network(tcsn)",this paper studies several technologies of train network fault detection and location,which has a certain engineering reference value for the realization of train network fault diagnosis and train intelligence.The main contents of this paper are as follows:(1)In this paper,the fault source and fault propagation model of train communication network is analyzed,the main theoretical methods and technologies of network fault detection and locationis deeply studied,and the causal diagram fault propagation model,bipartite graph fault propagation model and network tomography model is studied.The bipartite graph model and the network tomography model are both Boolean programming problems.Tomo algorithm and linear relaxation algorithm are studied.(2)Aiming at the problem that the connector of vehicle equipment falls off,the motion detection technology based on video is studied.The improved vibe motion detection algorithm combined with color threshold segmentation is studied and implemented,which solves the "ghost" problem of vibe algorithm,and realizes the detection of looseness fault of interface,a specific object.(3)The SNMP(Simple Network Management Protocol)protocol is deeply studied to detect the faults of the main control computer and other network equipment.The fault information collection of disk,CPU,process and other objects of vehicle equipment is realized by polling.A fault detection scheme based on SNMP is designed.(4)According to the fault propagation relationship of on-board equipment,the bipartite graph fault propagation model of central control unit is established.Tomo algorithm and linear relaxation algorithm are programmed to simulate the bipartite graph solution of central control unit.The simulation results show that the two algorithms can achieve high-precision fault detection of central control unit.(5)Fault medium location for train network.The network tomography models of the main train level and vehicle level networks are established.The linear relaxation algorithm and greedy heuristic algorithm are also used to solve the model.The network tomography models of WTB bus network and ARCNET double loop network are established on the train level network.The network layer of CRH5 power unit and CRH3 traction unit is established on the vehicle level network Analyze the imaging model.Through the fault medium location test of each network.The results show that: compared with tomo algorithm,linear relaxation algorithm has higher fault location accuracy in large model.In order to solve the problem that the computational efficiency of the linear relaxation algorithm is low and decreases with the scale of the model,the improved model is the packet network tomography model.Simulation results show that this method can ensure the accuracy of the linear relaxation algorithm and reduce the calculation time.